Research On The Logic Of Information Flow For Cognitive Metaphors

The perspective of cognitive metaphor no longer regards metaphor simply as a language skill but a basic way of human cognition.It is an explicit manifestation of human’s complex thinking mode.This tendency makes researchers focus not on the lexical level of metaphors,but on the semantic-pragmatic interface and the reasoning mechanism.With these studies,they try to reveal how humans acquire knowledge and transmit information.However,the cognitive metaphor theory highly relies on the intuition of linguists and lacks programmatic analysis.With the rapid development of natural language processing technology,metaphor recognition and metaphor interpretation also become a highly concerned topic.Currently,automatic processing of metaphors requires large-scale databases,and a generalized formal model can overcome the limitations of personal preference and specific corpus.Therefore,the formalization of metaphor understanding also become an important task.Based on situation theory and epistemic logic,this study explores the definition,characteristics and understanding process of cognitive metaphors from the perspective of information flow.By providing corresponding formalization methods,the study discusses the information transmission in metaphor judgment,understanding,and belief updating,revealing the general framework and specific properties of metaphorical cognition.This dissertation is divided into eight chapters,and the main contents of each chapter are arranged as follows:Chapter 1 is the introduction,which first describes the research background and outlines the history and current situation of the two main theoretical starting points of this study,namely the cognitive research and the formalization of metaphor.Then,it proposes that this study hopes to deepen the understanding of metaphorical reasoning process by logic representation of cognitive metaphor theory and inspire the automatic processing of metaphor.Finally,the framework structure of the dissertation is introduced.Chapter 2 briefly reviews definitions of metaphors in mainstream metaphor theories,and clarifies the basic standard of metaphor decision and the main properties of metaphorical expression.Then,starting from the truth conditions of metaphor,the generation of metaphorical semantics are explored.Since metaphorical statements need to generate non-literal meanings,they cannot fully follow the principle of semantic compositionality.We propose a solution by adding a process of semantic adaptation,which Recanati calls semantic modulation,to make the semantic involved in combination no longer literal meanings.And since the modulation function essentially reflects the influence of different contexts on semantics,the choice of the meaning of metaphors in specific contexts is dynamic.Therefore,following the tradition of dynamic semantic research,we take metaphorical semantics as an information change potential that guides the transformation of literal information into non-literal information,and analyze the transformation flow in this process.Chapter 3 introduces two main paths of information flow research: one treats information as range,for example,epistemic logic and doxastic logic;the other treats it as correlation,like situation theory and information channel theory;and the attempt to construct logics that fuses the two viewpoints.These classical ways of dealing with information flow is the background knowledge of this dissertation and the basis for the formalization of metaphors we show below.Chapter 4 and Chapter 5 discuss metaphor comprehension in static environments,that is,how to understand a metaphorical sentence itself.The Chapter 4 analyzes the information structure of metaphor comprehension based on situation theory and situation semantics,and discusses the concept of metaphors,the information structure in a single metaphoric expression,and how the literal meaning is transformed into non-literal information.Based on the analysis in Chapter 4,Chapter 5 firstly introduces a metaphor operator and a preference operator on the basis of minimal modal logic to construct a metaphor decision logic based on ”infons” to understands conceptual mapping at the level of information.Then a metaphor comprehension logic is given to make a unified interpretation of metaphors in both standard forms and special forms(e.g.negative metaphorical expressions,indirect metaphorical expressions,etc.).Also,we illustrate these characteristics through case studies.Chapter 6 and Chapter 7 discuss the metaphor comprehension process in a dynamic environment.In this part,we illustrate the influence of contextual elements such as presupposition,context,and discourse situation on the understanding of metaphorical sentences.Chapter 6 analyzes the dynamic information flow of metaphor comprehensions based on information channel theory.By regarding metaphor comprehensions as the process of transferring the source-related information to the target-related information,to explain how the information of different conceptual domains is mapped and integrated to form the information channel required for information transmission.And the constraints required for information transmission can be expressed.Based on the model in Chapter 6,Chapter 7 further considers belief revision of metaphor comprehension.Using dynamic doxastic logic language and spherical system semantics by adding consideration of the degree of reasonableness,the belief revision of metaphorical comprehensions can be formalized.Then with examples in ordinary expressions,we describe the special properties of the metaphorical information update,such as the information is not uniquely determined,the agent has a preference for the information,and the source domain and target domain information are mutually restricted.Finally,Chapter 8 summarizes the main conclusions and contributions of this dissertation,and pointing out its innovations.It also analyzes the existing deficiencies,and points out the orientation of future work.